Today, we explore the relationship between failure modes and catastrophic failures. Catastrophic failure of tailings dams and its water management structures are the likely result of the failure of a number of essential components or any combination of states. We purposely use different terminology instead of failure modes because they explain how the dam can fail, but not why.
If we look at past events, like the 2017 Oroville water dam failure and the 2015 Fundão tailings dam failure, we can see that failure often occurs due to a complex set of causes. If the methodology does not address combined states and failure modes, it will not depict a fair representation of the risks the dam generates. Indeed, analyses and the resulting evaluation of probabilities must be repeatable, traceable, defensible, and be supported by relevant data.
Specific Requirements for Dam Analysis
In order to address these dams’ specific requirements, it is important to consider all the analyses the engineers have carried out. These generally encompass static and pseudostatic factor of safety of the slope evaluated using undrained strength analysis, static effective stress analysis, and residual-liquefied strengths, as appropriate. Additionally, we include uncertainties/variability resulting from the initial geotechnical site characterization/investigations, construction type and materials, human error, ancillary water management facilities, monitoring and inspections and finally management, documentation of changes and maintenance actions in the assessment of the probability of failure.
Caveat
We recognize that in doing so, the risk analysis involves a large number of considerations. Not all of them are amenable to probabilistic modeling and analysis. The use of expert opinion (subject matter experts/peer review) in the risk analysis allows the inclusion of uncertainties that might otherwise be difficult to calculate or quantify.
Expert elicitation is one of several methods acceptable in USACE guidance documents for use in reliability analyses and risk assessments (SACOE ETL 1110-2-561, U.S. Army Corps of Engineers, Reliability Analysis and Risk Assessment for Seepage and Slope Stability Failure Modes for Embankment Dams, 2006).
Conclusion
Using the procedure outlined above, that we have crystallized in the ORE2_Tailings™ methodology, we ground dams’ risk assessments in defensible data. In addition, we incorporate engineering analyses model results, judgement and expert opinion, as well as benchmarks related to the historical database of failures. One can consider ORE2_Tailings a quantitative evaluation of the probability of failure and the consequences. We blend that evaluation with semi-empirical rules integrating all the aspects that cannot be quantified in a purely geomechanical approach.
Climate change is rapidly altering the environment. As a result, it is paramount to use tools that swiftly assess risks and allow users to make non-partisan risk-informed decisions. Those decisions must ensure geoethical practices in the form of a fair representation of the risks.
